Abstract

Data-driven models are becoming increasingly competent at tasks fundamental to weather and climate prediction. Relative to machine learning (ML) based atmospheric models, which have shown promise in short-term forecasting, ML-based ocean forecasting remains less explored. In this seminar we first present results from idealised experiments; training a UNet to emulate a GCM with a simplified channel model configuration. We show that interactions with land are especially challenging. This systematic bias highlights the need for careful and thorough model assessment, and emphasises one of the many differences between modelling the atmosphere and the ocean. We then present work developing an ocean component for a data-driven earth system model, based on the AIFS weather prediction model developed at ECMWF . We give an overview of the AIFS weather prediction system, and then highlight the various design considerations in developing a data-driven earth system model and show preliminary results from this work.